Debates over systems like Israel's 'Lavender' often focus on the AI. However, the more critical issue may be the human-defined 'rules of engagement'—specifically, what level of algorithmic confidence (e.g., 55% accuracy) leadership deems acceptable to authorize a strike. This is a policy problem, not just a technology one.

To prevent a scenario where 'the algorithm did it,' the U.S. military relies on the legal principle of 'human responsibility for the use of force.' This ensures a specific commander is always accountable for deploying any weapon, autonomous or not, sidestepping the accountability gap that worries AI ethicists.

The military's AI use is overwhelmingly focused on non-lethal applications like logistics and processing intelligence data. The 'pointy end' of autonomous weapons represents just one small category within a much broader AI strategy that mirrors corporate use cases.

Despite hype in areas like self-driving cars and medical diagnosis, AI has not replaced expert human judgment. Its most successful application is as a powerful assistant that augments human experts, who still make the final, critical decisions. This is a key distinction for scoping AI products.

Instead of automating decisions, the Pentagon's AI strategy focuses on synthesizing vast amounts of data—assets, weather, potential reactions—to expand a human operator's situational awareness, enabling them to make better, more informed choices.

Defense tech firm Smack Technologies clarifies the objective is not to remove humans entirely. Instead, AI should handle low-value tasks to free up personnel for critical, high-value decisions. This framework, 'intelligent autonomy,' orchestrates manned and unmanned systems while keeping humans in the loop.

Contrary to the 'killer robots' narrative, the military is cautious when integrating new AI. Because system failures can be lethal, testing and evaluation standards are far stricter than in the commercial sector. This conservatism is driven by warfighters who need tools to work flawlessly.

Countering the idea that slow, manual processes add valuable friction to warfare decisions, the Pentagon's view is that AI maintains critical checks and balances (rules of engagement, approvals). It only removes the inefficient friction of "hunting and pecking" for data, leading to faster and better-informed decisions.

In operations, AI models like Anthropic's Claude are used for intelligence analysis, summarizing media chatter, and running simulations to aid commanders. They are not used for autonomous targeting; any outputs go through layers of human review before influencing battlefield decisions.